Spoken digit classification using a spin-wave delay-line active-ring reservoir computing

TL;DR

This paper demonstrates the use of a spin-wave delay-line active-ring reservoir computer for spoken digit recognition, achieving high accuracy and showcasing its potential for tasks requiring short-term memory and nonlinearity.

Contribution

It introduces a novel spin-wave reservoir computing device with improved transducers and evaluates its performance on speech recognition and memory tasks.

Findings

Achieved up to 93% accuracy in spoken digit classification.

Demonstrated effective short-term memory and parity check capabilities.

Validated the device's potential for reservoir computing applications.

Abstract

As a test of general applicability, we use the recently proposed spin-wave delay line active-ring reservoir computer to perform the spoken digit recognition task. On this, classification accuracies of up to 93% are achieved. The tested device prototype employs improved spin wave transducers (antennas). Therefore, in addition, we also let the computer complete the short-term memory (STM) task and the parity check (PC) tasks, because the fading memory and nonlinearity are essential to reservoir computing performance. The resulting STM and PC capacities reach maximum values of 4.77 and 1.47 respectively.